11651 2017-06-13 09:08:14 +0000 test_sstate_noop_samesigs performance when there are differences is dire 2017-08-16 19:29:48 +0000 1 1 10 QA/Testing Build Testing general unspecified x86 Multiple RESOLVED FIXED Medium normal 2.4 1 richard.purdie leonardo.sandoval.gonzalez anibal.limon --- --- No (bug/feature does not impact docs) oldest_to_newest 74044 0 richard.purdie 2017-06-13 09:08:14 +0000 oe-selftest -r sstatetests.SStateTests.test_sstate_noop_samesigs I had a local issue where a function depended on BB_NUMBER_THREADS in global scope (in do_package_ipk). This selftest took hours to run. The reasons were that: a) The test forces single threading (easy fix, I'll send a patch) b) It loops over calling bitbake-diffsigs on every hash difference thanks to: http://git.yoctoproject.org/cgit.cgi/poky/commit/meta/lib/oeqa/selftest/sstatetests.py?id=85dbd7bf9e4b6eccdbb8102e462dd681240c1c57 Whilst bitbake-diffsigs is useful we probably need to cap the output at X (say 20?) differences. 74045 1 richard.purdie 2017-06-13 09:12:37 +0000 One quick fix may be to call the python function instead of using process and a new python interpreter each time. 75660 2 leonardo.sandoval.gonzalez 2017-08-04 19:01:44 +0000 patches sent to the mailing list: http://lists.openembedded.org/pipermail/openembedded-core/2017-August/140480.html and 81. 75919 3 leonardo.sandoval.gonzalez 2017-08-16 19:29:48 +0000 Merged at a2e2f434cd: Leonardo Sandoval: Fri Aug 11 00:08:33 2017 +0100: sstatetests: limit the number of signature comparisons when differ 5dd65cf502: Leonardo Sandoval: Tue Aug 8 12:53:02 2017 +0100: sstatetests: Use python function instead of bitbake-diffsigs script